+ All Categories
Home > Documents > An Investigation of the Performance of Hot Mix Asphalt ... An Investigation of the Performance of...

An Investigation of the Performance of Hot Mix Asphalt ... An Investigation of the Performance of...

Date post: 02-Apr-2018
Category:
Upload: lethuan
View: 221 times
Download: 1 times
Share this document with a friend
69
1 An Investigation of the Performance of Hot Mix Asphalt (HMA) Binder Course Materials with High Percentage of Reclaimed Asphalt Pavement (RAP) and Rejuvenators By Ram Kumar Veeraragavan A Thesis Submitted to the Faculty of the Worcester Polytechnic Institute in partial fulfillment of the requirements for the Degree of Master of Science in Civil Engineering ______________________________________ April 2016 Approved _____________________ Dr. Rajib B. Mallick _____________________ Dr. Mingjiang Tao _____________________ Dr. Tahar El-Korchi
Transcript

1

An Investigation of the Performance of Hot Mix Asphalt (HMA) Binder Course Materials

with High Percentage of Reclaimed Asphalt Pavement (RAP) and Rejuvenators

By

Ram Kumar Veeraragavan

A Thesis

Submitted to the Faculty

of the

Worcester Polytechnic Institute

in partial fulfillment of the requirements for the

Degree of Master of Science

in

Civil Engineering

______________________________________

April 2016

Approved

_____________________

Dr. Rajib B. Mallick

_____________________

Dr. Mingjiang Tao

_____________________

Dr. Tahar El-Korchi

i

Abstract

Use of high percentage of Reclaimed Asphalt Pavement (RAP) material in Hot Mix Asphalt is one

of the several sustainable construction alternatives being considered by many Departments of

Transportation (DOT). Use of RAP in HMA helps in reducing the consumption of virgin aggregates

and binder and construction cost and conserving energy. Although most of the US state agencies

allow the use of 30% or more RAP in the design of Hot Mix Asphalt (HMA), the current average

RAP usage is only about 10 to 20%. This is because of the uncertainty about the performance of

HMA mixes with a high RAP content. Several factors influence the performance of the HMA mixes

with a high RAP content. Recent research has shown that the use of a high RAP content in HMA

with rejuvenators is successful in reducing the stiffness of the RAP mixes, and thereby improving

their performance. The present work is carried out to explore the feasibility of using a high RAP

content of 50% in a binder layer HMA with the addition of rejuvenators.

Ultrasonic Pulse Velocity (UPV) test was carried out to compare the stiffness of the RAP

mixes with and without the addition of rejuvenators. Moisture Induced Stress Test (MIST) was

conducted to study the effect of moisture damage on the HMA mixes with high RAP content. The

Indirect Tensile Strength Test (ITS) was used to determine the strength of the HMA mixes with high

RAP content. In addition, creep compliance and Semicircular Bend (SCB) tests were carried out to

determine the cracking potential and fracture strength of the mixes respectively. The addition of

rejuvenators was found to significantly reduce the stiffness of the mix with high RAP content. The

predicted complex shear modulus (G*) obtained from the Hirsch model and performance grading

tests on extracted binders confirmed the effectiveness of the addition of rejuvenators in reducing the

stiffness of recycled asphalt binder in the recycled mixes.

ii

Acknowledgements

I sincerely thank Professor Rajib Mallick for giving an opportunity to study at Worcester Polytechnic

Institute and providing me with necessary funding and support. I am grateful for his continuous

encouragement and support throughout the project work. I would also like to thank the members of

my thesis committee, Prof. Tahar El-Korchi and Prof.Mingjiang Tao for their valuable comments

on my research work.

I am thankful to the laboratory managers, Don Pellegrino and Russ Lang for their co-operation and

help in the operation of the various laboratory equipment and resolving problems with the working

of the various equipment. I would also like to specially thank my colleagues Uma Maheswar Arepalli

and Sakthivelan Ramachandran for their help and support during the project work and analysis of

the data. The great number of testing works were possible only through the help from my fellow

graduate colleagues Ryan Worsman and Rachel Kennedy. I would also like to thank the Maine

Department of Transportation (DOT) for providing all the needed resources for this project work.

Last but not least, I would like to thank my parents for their continuous support and encouragement

throughout my project work.

iii

TABLE OF CONTENT

Abstract…………………………………………………………………………………….…..i

Acknowledgements……………………………………………………………………….…...ii

TABLE OF CONTENT..……………………………………………………………………...iii

LIST OF TABLES……………………………………………………….…………...…..…...v

LIST OF FIGURES...……………………………………………………………..….……….vii

1 INTRODUCTION ...................................................................................................................... 1

Use of high RAP content in hot mix asphalt ....................................................................... 1

Factors influencing the characterization of RAP ................................................................ 2

Use of rejuvenators to improve the properties of RAP mixes............................................. 2

Objectives ............................................................................................................................ 4

Scope of work...................................................................................................................... 4

2 LITERATURE REVIEW ........................................................................................................... 5

Importance and need for conservation of natural resources ................................................ 5

Needed research ................................................................................................................ 13

3 EXPERIMENTAL INVESTIGATIONS ................................................................................. 14

Experimental plan ............................................................................................................. 14

4 MATERIALS AND METHODS ............................................................................................. 16

Gradation ........................................................................................................................... 16

Virgin Binder..................................................................................................................... 17

Rejuvenators ...................................................................................................................... 17

Waste vegetable oil (WV oil) ............................................................................................ 17

Sylvaroad (SV) .................................................................................................................. 18

5 MIX PROCEDURE .................................................................................................................. 19

20% RAP and 50 % RAP without rejuvenators ................................................................ 19

50% RAP with waste vegetable oil rejuvenator ................................................................ 19

RAP with rejuvenator Sylvaroad (SV) .............................................................................. 20

6 COMPACTION ........................................................................................................................ 21

7 TEST METHODS .................................................................................................................... 22

Ultrasonic pulse velocity ................................................................................................... 22

iv

Creep compliance and Indirect Tensile Strength test (ITS) .............................................. 23

Indirect Tensile Strength (ITS) ......................................................................................... 25

Moisture Induced Stress Tester (MIST) ............................................................................ 26

Semi Circular Bend test (SCB) ......................................................................................... 26

Asphalt binder extraction .................................................................................................. 28

8 HIRSCH MODEL .................................................................................................................... 29

9 RESULTS AND ANALYSIS .................................................................................................. 30

Ultrasonic Pulse Velocity (UPV) test result ...................................................................... 30

MIST test results ............................................................................................................... 34

Penetration test results ....................................................................................................... 38

Performance Grade (PG) results: ...................................................................................... 40

Back calculated Hirsch model results ............................................................................... 41

Indirect Tensile Strength (ITS) results .............................................................................. 43

Creep compliance results .................................................................................................. 45

Semi Circular Bend (SCB) test results .............................................................................. 47

Cost analysis ...................................................................................................................... 52

Summary of tests ............................................................................................................... 53

10 CONCLUSIONS................................................................................................................... 55

11 RECOMMENDATIONS ...................................................................................................... 56

12 REFERENCES: .................................................................................................................... 57

v

LIST OF TABLES

Table 1. Gradation of Materials Used in the Present Study…………………………………………16

Table 2. Individual aggregate proportions and asphalt content………………………………....…..16

Table 3. Saturates, Asphaltenes, Resins, Aromatics (SARA) table for Waste Vegetable Oil……….18

Table 4. Virgin and RAP Binder content in HMA Mixes…………………………………………...20

Table 5. Waste Vegetable Oil rejuvenator content calculation……………………………………..21

Table 6. Sylvaroad Rejuvenator Content Calculation…………………………………………..…..21

Table 7. Air Void Distribution among samples…………………………………………………….31

Table 8. Seismic and Design Moduli at 25°C and -10°C……………………………………………32

Table 9. ANOVA for Design Modulus (DM) at 25°C……………………………………………...34

Table 10. ANOVA for Design Modulus (DM) at -10°C……………………………………………35

Table 11. Moisture Induced Sensitivity Tester (MIST) results on Moduli Values of RAP mixes......36

Table 1: Density comparison (PreMIST vs PostMIST)………………………………………..…..37

Table 13. ANOVA for Pre-MIST DM at 25°C………………………………………………….….38

Table 14. ANOVA for Post-MIST DM at 25°C…………………………………………………….38

Table 15. Penetration Test Results on Extracted Binder from HMA mixes………………………...39

Table 16. ANOVA for penetration at 25°C…………………………………………………….…..40

Table 17. Performance Grade (PG) -Low and High results…………………………………………41

Table 18. Volumetric properties……………………………………………………………………42

Table 19. Back calculated G* using Hirsch Model…………………………………………………43

Table 20. ANOVA for back calculated G* at 25°C ………………………………………………..43

Table 21. Ranking of Mix types……………………………………………………………….........44

Table 22. Indirect Tensile Strength Test (ITS) Results……………………………………………..44

Table 23. ANOVA for ITS at 25°C ………………………………………………………………..46

Table 24. Ranking for ITS at 25°C……………………………………………………………….....46

Table 25. Creep compliance test results……………………………………………….…………....47

Table 26. Fracture Energy test results of 20% RAP ………………………………………………..49

vi

Table 27. Fracture Energy test results of 50% RAP ……………………………………………......49

Table 28. ANOVA for fracture energy at 25°C for 20% RAP………………………………………52

Table 29. ANOVA for fracture energy at 25°C for 50% RAP…………………………………........52

Table 30: Cost of materials per ton…………………………………………………………………53

Table 31: Density of Rejuvenators…………………………………………………………………53

Table 32. Cost Analysis………………………………………………………………………….....54

vii

LIST OF FIGURES

Figure 1. Experimental plan..............................................................................................................14

Figure 2. Optimum rejuvenator dosage for Waste Vegetable Oil……..…………………………....18

Figure 3. (a)Superpave Gyratory compactor and (b) Saw cutter………………..…………………..22

Figure 4. Ultrasonic device…………………………………………………….……...……….…...24

Figure 5. Creep compliance test setup………………………………………….……….………….26

Figure 6. Moisture Induced Stress Test…………………………………………..……….………...27

Figure 7. Semicircular bend test (SCB) and specimen…………………………..……….…………29

Figure 8. Average Design Modulus…………………………………………….…………..………33

Figure 9. Plots of temperature versus modulus for the different mixes………..………..…………34

Figure 10. Comparison of Pre, Post MIST design Moduli and Air voids………………………….37

Figure 11. Penetration results…………………………………………………….………….……..40

Figure 12. PG- Low High………………………………………………………….………….……41

Figure 13. Indirect Tensile strength results………………………………………….…………..….45

Figure 14. Creep compliance at -10C……………………………………………….……………...47

Figure 15. Fracture Energy at different rejuvenator dose for A) 20% RAP B) 50% RAP………......50

Figure 16. Comparison of Fracture Energy of 20% RAP and 50% RAP mixes …...……………......51

Figure 17. Comparison of Flexibility Index of 20% RAP and 50% RAP mixes …………………..51

Figure 18. Radar chart of different test results……………………………………………….…….55

1

1 INTRODUCTION

One of the most difficult challenges for the development of any road network is to execute projects

in harmony with the concept of sustainable development. The road industry is therefore looking

forward for alternative materials and construction technology, which are environment friendly,

energy efficient and cost effective for the construction and maintenance of roads. Most of the

current road construction practices are primarily dependent on naturally occurring aggregates that

are obtained from quarries. The extraction of these aggregates from their natural sources results in

the loss of forest cover and pollution on a large scale leading to environmental degradation. This,

in turn, has raised environmental concerns in many parts of the world [1]. In order to sustain natural

resources, sufficient reserves have to be ensured to meet the demands of aggregates at present and

in the future, as these resources are depleting fast and are non-replenishable. On the other hand,

the price of asphalt binder has also been fluctuating, and research is needed to reduce the

consumption of virgin asphalt binder in rehabilitation strategies through alternate technologies and

thereby reduce the cost of construction and maintenance. To cope up with the demand for

aggregates to preserve and maintain the road infrastructure assets, many Departments of

Transportation (DOT) are using Reclaimed Asphalt Pavement (RAP) material as an alternative [2],

and considering their use at an increased percentage for Hot Mix Asphalt (HMA). Recycling of

asphalt pavement materials has proved to be a valuable approach for both economic and

environmental reasons.

Use of high RAP content in hot mix asphalt

The use of RAP in road construction can provide savings from 14% to 34% for RAP percentages

varying from 10% to 50% of the total mix, [3]. Although it is generally accepted that the utilization

of RAP in HMA can reduce cost, the percentage of RAP to be used in conventional HMA mixes

is still debated. When using high RAP content in HMA, it is still unclear as how the aged asphalt

binder from RAP interacts with virgin asphalt binder. It is generally assumed that the RAP material

will not act like a black rock in the new mix and it will blend well with virgin binder during mixing

[4] , but the question remains as to what extent will it blend. A constant effort is being made by the

DOTs to adopt a higher RAP percentage for regular HMA production so as to reduce the

consumption of virgin aggregates, cost and conserve energy.

2

Factors influencing the characterization of RAP

RAP is milled pavement material obtained from old and often distressed asphalt concrete

pavement. This RAP material mainly consists of asphalt binder and aggregates. Of these two

materials, the asphalt binder generally undergoes various physical and rheological changes during

their service life. The binder properties in RAP are significantly affected by the rheological

changes. There are two predominant factors which account for the severity of change in RAP

binder properties: composition of the original binder used during its construction and amount of

aging undergone during its service [5]. The asphalt binder in RAP, during its service period

undergoes two stages of aging, i.e. short term aging (during construction) and long term aging

(during service). During these two stages of aging, the viscosity of the binder gradually increases

due to natural phenomena like oxidation, volatilization, polymerization and syneresis making it a

hard and a stiff material [6]. This aged RAP asphalt binder significantly influences the property of

the new blend.

Use of rejuvenators to improve the properties of RAP mixes

The four components of asphalt are: Saturates, Aromatics, Resins, and Asphaltenes (SARA) [7].

The asphalt in HMA mixes undergoes aging during the manufacturing process and this is known

as short term aging. The binder in HMA layers when exposed to the atmosphere during the service

life, undergo long term aging. During the aging process, the percentage of asphaltenes in the binder

increases in proportion to the other components due to volatilization and oxidation. Higher

percentage of asphaltenes in the binder stiffens the binder, causing cracks in the HMA layers.

Therefore, the transformation in the components in the asphalt binder, causes stiffening of the

HMA mixes and leads to cracking. The properties of the binder as obtained from the refinery is

very different from the aged binder in the field due to short and long term aging.

Several factors have been identified as potential causes of aging [8] and they are as follows:

1. Oxidation; 2. Volatilization; 3. Steric or physical hardening; 4. Exudation of oils; 5. Photo-

oxidation by direct light; 6. Photo-oxidation in reflected light; 7. Photochemical reaction

3

by direct light; 8. Photochemical reaction by reflected light; 9. Polymerization; 10. Changes

by nuclear energy; 11. Action of water; 12. Absorption by solids; 13. Absorption of

components at a solid surface; 14. Chemical reactions; 15. Microbiological deterioration.

The following four factors are considered to be the most significant factors that contribute to the

aging process.

1. Oxidation: Due to oxidation, the components of the asphalt binder oxidize. They form heavier

and more complex molecules. This results in increased stiffness and decreased flexibility. It is to

be noted that the rate of oxidation is significantly affected by the temperature and the asphalt binder

film thickness. Thinner films in HMA mixes when exposed to high temperature enhances

oxidation. It is reported that at temperatures above 100°C, the rate of oxidation doubles for every

10°C increase in temperature.

2. Volatilization: Due to volatilization of the lighter components in the binder, the binder stiffens

and gets aged.

3. Steric or physical hardening: The steric or physical hardening occurs due to reorientation of

molecules and slow crystallization of waxes contributing to hardening of the binder, even at

ambient temperatures.

4. Exudation of oils: Some of the aggregates are porous. Due to the porous nature of the aggregates,

the oils from the asphalt binder are exuded into the aggregates in an asphalt mix to a different

extent thereby resulting in aging of the binder.

Rejuvenating agents are added to RAP mixes to restore the physical and chemical

properties of the aged RAP binder. Rejuvenating agents are additives, which are capable of

restoring the original rheological properties of the aged RAP binder [9].These rejuvenating agents

are believed to diffuse through the aged RAP binder up to a certain depth of the aged binder film

and restore the original maltene to asphaltene ratio in it, making it a less stiff material or a more

flexible material [10].This rejuvenated RAP binder is expected to blend with the virgin binder and

contribute to the overall asphalt content of the mix. Recent research has shown that the use of a

high (90% - 100%) RAP content in HMA with rejuvenators is successful in reducing the overall

4

stiffness of the RAP mixes, and improving its performance. But research on the use of rejuvenators

and its effects on performance of mid-high range RAP mixes has been so far very limited.

Objectives

The objective of this study is to investigate the performance of hot mix asphalt (HMA) binder

course materials with 50% reclaimed asphalt pavement (RAP) with rejuvenators and compare test

results with that of a control mix with 20 % RAP and to quantify the benefits of using HMA with

50% RAP content.

Scope of work

The scope of work consisted of preparation of HMA mixes with and without rejuvenators. Two

rejuvenators, Waste Vegetable Oil and Sylvaroad were used in the present investigation. The

various tests that were conducted on the HMA samples include non-destructive test (Ultrasonic

pulse velocity), creep test, indirect tensile strength and semicircular bend test. The mixes were also

subjected to Moisture Induced Stress Testing (MIST) to evaluate their moisture susceptibility. The

extracted binders were tested for their stiffness at low, intermediate and high temperatures to

evaluate their thermal, fatigue cracking and rutting potential respectively and determine the PG

grade. Finally cost analysis was carried out to compute the savings in total construction cost, if

HMA mixes with high RAP content are used.

5

2 LITERATURE REVIEW

Importance and need for conservation of natural resources

A literature review was carried out to determine the state of the art and practice for use of RAP in

HMA, its properties and performance with addition of rejuvenators, and its susceptibility to

moisture damage. The focus of the review was on the high percentage RAP use with rejuvenators

in binder and surface HMA layers.

Recycling of RAP has proved to be one of the most cost effective solutions to help manage the

rising cost of materials and increased demand of new aggregates. All roads need to be periodically

replaced or repaired and this work would produce reclaimed materials, which have considerable

value that can be reused. According to the National Asphalt Pavement Association [11] more than

500 million tons of reclaimed asphalt pavement are being produced each year in the US from

milling or other breaking up of old surfaced roads, of which only 100 million tons are being re-

used in pavement-related applications. This RAP material has good quality of mineral and filler

material and they might have been unchanged in properties over the years except for the binder

which would have age-hardened. It has been found that the most economical use of the RAP

material is in the surface and intermediate layer of HMA pavements, where the binder from RAP

can replace a portion of the more expensive virgin binder [12]. Even though RAP has been used for

road construction since the 1930’s in small percentages, there is a recent interest in using higher

RAP content in mix design. Recent research has shown that recycling with 90% - 100% RAP

content is possible with the addition of appropriate rejuvenators [2].

Milling operations generally result in production of fines in the RAP, which, from most wearing

surface mixes, usually have 4.5-6% asphalt content. RAP materials generally contain aged binder

with high stiffness, and as a result are believed to have inferior fatigue and thermal cracking

properties [13].

Therefore, in order to utilize RAP in HMA, it is essential to characterize the aggregate and asphalt

in RAP. The major factors that determine the final percentage of RAP to be used in HMA are

mixture properties, aggregate requirements, RAP handling and homogeneity, and project

economics [14]. It is very crucial to ensure that the RAP binder is capable of blending well with the

6

virgin binder and that the final blend would meet all the binder requirement. The aggregates in

RAP sometimes have some serious effect on the total mixture volumetric and performance.

Therefore it is necessary to take into account the design aggregate structure, crushed coarse

aggregate content, dust proportion and fine aggregate angularity of RAP aggregates. When the

aged binder from RAP is combined with virgin binder, the resultant binder grade is affected [15].

While at low RAP percentages, this change in binder grade is negligible, at high RAP percentages

the effect of RAP binder becomes significant.

Rejuvenators are chemical additives that are capable of restoring the physical and chemical

characteristics of aged asphalt binder in RAP.

Chen et al., (2015) [16] studied the application of rejuvenators and soft asphalt cements as recycling

agents in RAP mixes. Various dosages of recycling agents were added to aged binders recovered

from field samples. It is found that the performance of hardened binders can be improved

significantly with the addition of rejuvenators. The blends mixed with rejuvenators behaved better

under fatigue than those with softening agents. The large molecular size was shown to be a

characteristic of an asphalt blend. It is suggested that the changes in the carbonyl area may be used

to estimate the viscosity value of the asphalt blend. It was reported that the formation of the

carbonyl area in aged asphalt was reduced by adding recycling agents, which in turn changed the

physical properties of the blended binder in a predictable manner. Chen et al (2015) also developed

a model to detect the content of recycling agents.

Shen et al., (2007) [17] studied Superpave mixtures containing RAP with rejuvenating agents

including a rejuvenator and a softer binder. They also carried out indirect tensile strength and

evaluated the mixtures for rutting using the asphalt pavement analyzer. The results indicated, for

the mixtures tested for this project, that the properties of the recycled mixtures using the

rejuvenator were better than those containing the softer binder and that 10% more RAP could be

incorporated in the Superpave mixtures by using the rejuvenator than using the softer binder. It

was also reported that the blending charts established under the Superpave binder specifications

can be used to determine the content of the rejuvenator for recycling.

Li et al., (2014) [18] studied the influence of aged modified asphalt in reclaimed asphalt pavement

(RAP) mix. Styrene–butadiene rubber (SBR) latex, a polymer emulsion was used to blend

7

modified asphalt with conventional asphalt and it was found that SBR latex enhances the low-

temperature properties of RAP binder efficiently without causing any observable negative

influence on RAP mix compaction. SBR latex was found to improve the viscoelastic

characteristics and other performance of RAP mix, including the resistance to low-temperature

cracking, rutting, and moisture damage.

Bennert et al., (2014) [19] evaluated the Plant-Produced High-Percentage RAP Mixtures in the

Northeast. Three potential strategies were evaluated for incorporating higher RAP contents, using

a softer asphalt binder grade to offset the stiff RAP asphalt binder, limiting the amount of RAP

binder credited to the total asphalt content of the asphalt mixture, and using a performance based

specification for the high-RAP content mixture. If a softer binder is specified, the availability of

the softer binder and the cost implications are to be considered. A marginal improvement in low-

temperature cracking properties were observed, when softer grade of binder was used. Softer grade

of binder did not improve the intermediate temperature cracking performance. It was reported that

75% and 50% RAP mixtures achieved better intermediate fatigue performance when compared

with the baseline 100% RAP mixture as shown in the overlay tester and flexural fatigue test.

Cooper et al., (2015) [20] studied the use of recycled asphalt shingles (RAS) as a partial

replacement for aggregates and petroleum-based virgin asphalt cement binder. It is reported that

5% RAS without recycling agents had similar performance compared with the control asphalt

mixture containing no RAS at high, intermediate, and low temperatures. The inclusion of RAS

with and without recycling agents showed an improvement in rutting performance with no adverse

effect on moisture sensitivity compared with the control mixture without RAS. It is interesting to

note that as use of recycling agents was increased, the recycled binder ratio, and the intermediate-

and low-temperature performances of the mixture were adversely affected.

Mogawer et al., (2016) [21] used softer binders to compensate the stiffness of reclaimed asphalt

pavement (RAP) binders in mixtures. The effect of five asphalt rejuvenators on the performance

of a 50% RAP surface-layer mixture was evaluated relative to rutting and cracking. It was found

that the rejuvenators degraded the rutting resistance of the 50% RAP mixture, although the use of

PMA binders remedied these degradations. The rejuvenators were found to improve the fatigue

cracking resistance of the 50% RAP mixture to a level higher than that of all-virgin control mixture

and also the 50% RAP mixture with softer binder. It was concluded that a combination of an

8

asphalt rejuvenator and a PMA binder was required to yield a high RAP mixture with similar or

better performance than a similar conventional mixture.

Diefenderfer et al., (2016) [22] reported the dynamic modulus of field-produced and field-cured

recycled pavement materials from 24 projects constructed in the United States and Canada. It was

found that the binder from the existing reclaimed asphalt pavement may play a role in their stiffness

properties. The authors reported that the master curves showed that the use of chemical additives

increased the stiffness and reduced the temperature dependency of the recycled materials. The

master curves showed that the dynamic modulus values were similar when emulsified asphalt and

foamed asphalt were used as the stabilizing and recycling agents.

Ding et al., (2016) [23] studied the effect of incorporation of recycled (aged) binder into virgin

asphalt especially the mechanism of the diffusion process between virgin and aged binders.

Molecular dynamics (MD) simulation was employed to investigate the diffusion between virgin

and aged binders, while increasing the asphaltenes ratio on the basis of virgin binder. It was found

that the diffusion of large molecules in asphalt was a critical factor for the diffusion of binders and

that it was more susceptible to the changes of temperature. It was reported that adding rejuvenator

into aged binder could accelerate the inter-diffusion rate between virgin and aged binder to

maximum level and increase the efficiency of recycling.

Shen et al., (2007) [24] studied the performance-based properties of rejuvenated aged asphalt

binders containing a rejuvenator at various percentages, under high, intermediate and low

temperatures. The rejuvenator was found to affect the performance-based properties of both the

rejuvenated aged binders and the mixtures containing the rejuvenated aged binders significantly.

The properties of the asphalt paving mixtures with the rejuvenated binders were found to have

improved or at the same level as the properties of the virgin mixtures.

Zaumanis et al., (2014) [2] studied the feasibility of producing 100% recycled mixtures. The mix

design procedures and the best RAP management strategies were reported. A cradle-to-gate

analysis of environmental effects was presented and it was shown that 18 kg or 35% CO2 eq

savings per ton of produced 100% recycled asphalt mixture is possible when compared to virgin

mix, while cost analysis showed at least 50% savings in material related expenses.

9

Im et al., (2014) [25] studied the impacts of various rejuvenators on the performance and

engineering properties of hot-mix asphalt (HMA) mixtures containing recycling materials (i.e.,

RAP and RAS). They found that the use of rejuvenators improved cracking resistance of the

recycled mixes. They also found that incorporation of rejuvenators in the recycled materials

improved their moisture susceptibility and rutting resistance. They concluded that the performance

of the rejuvenators depend on degree of blending between the binder of the recycled materials and

the virgin binder, aggregates, and the rejuvenator dosage.

Zaumanis et al., (2014) [26] reported the changes in Superpave performance grade (PG) of

Reclaimed Asphalt Pavement (RAP) binder after addition of two doses of six rejuvenators. The

high and low PG temperatures were found to reduce linearly with an increased dose while the

penetration value was found to increase exponentially. It was found that the grade sum of

rejuvenated RAP binder is always higher than that of the corresponding virgin binder.

Zaumanis et al., (2014) [27] carried out experiments with 100% recycled HMA laboratory samples

with five generic and one proprietary rejuvenators at 12% dose and studied the binder and mixture

properties. Waste Vegetable Oil, Waste Vegetable Grease, Organic Oil, Distilled Tall Oil, and

Aromatic Extract were found to change the Superpave performance grade (PG) from 94–12 of

extracted binder to PG 64-22 at similar doses while waste engine oil required higher dose to do

the same. The mixes with all rejuvenators ensured excellent rutting resistance while providing

longer fatigue life when compared to virgin mixtures and most lowered the critical cracking

temperature. It was found that rejuvenated samples required more gyratory compaction to reach

the design density compared to virgin samples and some oils reduced moisture resistance slightly.

Buss et al., (2015) [28] studied the rheological effects of warm mix asphalt (WMA) additives in

RAP mixes. They explored if the reduction in the asphalt binder grade is still detectable after in-

service aging. It was found that WMA facilitates the incorporation of higher amounts of recycled

asphalt materials. The recycled binder was found to have a larger influence on binder properties

compared to WMA additives.

Ongel and Hugener., (2015) [29] studied the potential reductions in construction costs and

environmental emissions by the use of reclaimed asphalt pavement (RAP) as aggregates in HMA

as an attractive alternative to the highway agencies. The aging behavior of rejuvenated 100% RAP

binder was studied and compared with that of the virgin bitumen. Three types of rejuvenators were

10

assessed in the study. It was found that the laboratory aging of a 70/100 Pen graded asphalt was

slower than that of rejuvenated 40/50 Pen grade asphalt. No significant difference between the

aging behaviors of the bitumen mixed with different types of rejuvenators was reported.

Xiao et al., (2015) [30] explored the rheological properties of the high percentage (up to 50%) of

RAP binder with three base binders in terms of five RAP binder content, two RAP binder sources,

one HMA and one warm mix additive (WMA) technologies at three aging states. The viscosity,

failure temperature, rutting resistance, fatigue resistance of various asphalt binders were tested.

Increasing the RAP binder concentration was found to improve the rutting resistance of asphalt

binder but a reduction in the fatigue resistance was noticed. RAP source was found to affect the

performance properties of combined asphalt binder.

Nazzal et al., (2015) [31] adopted Atomic Force Microscopy (AFM) techniques to study the effects

of rejuvenators on the nanomechanical properties of the interfacial blending zone that forms

between RAP and virgin asphalt binders in a high RAP content mixture. It was found that that the

rejuvenators did not have a significant effect on the modulus of the virgin binder. However, the

indentation modulus of the interface blending zones was found to have significantly decreased.

The authors reported that AFM force spectroscopy results showed that the rejuvenators increased

the interfacial blending zone adhesive bonding energy. The AFM indentation modulus of

interfacial blending zone was found to be correlating well with the Hamburg Wheel Tracking test

results of the high RAP content mixtures. The authors concluded that the interfacial blending zone

bonding energy might be one of the factors dictating fatigue performance of high RAP mixtures.

Ma et al., (2015) [32] investigated the feasibility of using high-content reclaimed asphalt pavement

(RAP) in high modulus asphalt concrete (HMAC). The effects of RAP content on performance of

recycled high modulus asphalt concrete (RHMAC) is found to be dependent on the specific RAP

content and the performance indicator. RAP content was found to have a significant influence on

dynamic modulus and failure strain when the RAP content increased to 40%, while the RAP

content showed significant impact on dynamic stability and tensile strength ratio when the RAP

content increased to 50%. Considering the influences of RAP on low temperature performance

and moisture stability, the authors suggest that RHMAC is not to be used in the surface layer with

high RAP contents.

11

Ali et al., (2016) [33] investigated the ability of five asphalt binder rejuvenators to restore low and

high temperature true performance grades of aged binders. The rejuvenators considered were:

Naphthenic Oil, a Paraffinic Oil, an Aromatic Extracts, a Tall Oil, and an Oleic Acid. Several sets

of asphalt mixtures containing different percentages (i.e. 25% and 45%) of RAP materials were

prepared using PG 76-22 polymer-modified asphalt binder and blended with rejuvenators at

manufacturers’ recommended dosage. The authors concluded that rejuvenators helped in lowering

the true grade of aged asphalt binders of RAP and that all of the rejuvenated binders had lower

performance grade than that of the control binder. The rejuvenator’s effectiveness was found to be

not affected by aging (from 2 to 6 h) and by increasing the amount of RAP materials (up to 45%).

The rejuvenators were also found to improve the fatigue resistance without substantially

influencing rutting performance.

Lu and Saleh., (2016) [34] investigated the performance of WMA with RAP at different

percentages, from 0 to 70% by mass of WMA. The performance of mixtures was compared with

a control HMA. Mixtures with the chemical additive were found to perform better than other

mixtures in terms of moisture resistance. WMA mixture with the rejuvenator showed a higher

number of cycles to fatigue failure than the control HMA. The increase in RAP proportion was

found to greatly improve the performance of rutting performance of WMA mixtures. All WMA-

RAP mixtures were found to offer better rutting resistance than the HMA.

Moghaddam and Baaj (2016) [35] presented an overview of the potential of using RAP with

rejuvenators. They concluded that the rejuvenating mechanism needs to be explored. The binder

and mixture performance tests can to some extent provide the behavior of rejuvenated mixture.

However, more advanced testing such as chemical tests should be performed to evaluate the

chemical performance characteristics of the rejuvenated binders.

Yu et al. (2014) [36] have conducted research on rheological properties of virgin, aged and

rejuvenator binders using dynamic shear rheometer and bending beam rheometer, and showed that

the viscosity and complex modulus of the rejuvenated binder, were between those of virgin and

aged binder.

Rojas et al., (1999) [37] evaluated asphalt mixes in the laboratory using ultrasonic pulse velocity

tests and concluded that the seismic modulus increased with a decrease in the voids in the total

mix (VTM). The seismic modulus decreased with a decrease in the binder viscosity; however, the

12

impact of the viscosity was found to become less pronounced at higher void percentage (as close

to 8% compared to 4% voids). Also at higher void levels the impact of binder viscosity on the

seismic modulus would be significantly high. The UPV is a suitable test as it allows the testing of

a sample before and after moisture conditioning due to its nondestructive nature.

The Semi-Circular Bending (SCB) test is a particularly attractive test because it uses a semicircular

sample – half of a standard laboratory compacted HMA sample or field core (and hence two test

samples could be obtained from one compacted sample or field core). Baoshan et al. (2005)

conducted Semi-Circular Bend (SCB) test and compared its results with Indirect Tensile Strength

(ITS) of the HMA mixes. Their research work concluded that SCB test, similar to ITS, can be used

to characterize the tensile strength of asphalt mixtures with good repeatability. The results from

their study showed that the tensile strengths of the SCB test and ITS were different because of

difference in stress states under loading. They also concluded that SCB was a more suitable test

for evaluating the tensile properties of HMA mixtures, because of smaller permanent deformation

under the loading strip.

Kakar et al., (2015) [38] presented a review of various techniques and investigations for assessing

the moisture damage so as to optimize the standard testing protocols. The authors concluded that

introduction of new in-situ testing techniques and material selection criteria is required to address

the moisture susceptibility of asphalt mixtures, which can improve the field assessment of moisture

damage that appears during the design life of an asphalt pavement, and bridge the gap between

field and laboratory investigations.

Tran et.al., (2012) [39] evaluated the effect of using a rejuvenator on mechanistic and performance

properties of recycled binders and mixtures containing high RAP and RAS contents in the

laboratory. Their research finding showed that the use of rejuvenator in the recycled mixtures

improved the cracking resistance of the RAP mixtures without severely affecting their resistance

to moisture damage and permanent deformation.

Willis et al., (2012) [40] conducted research to find whether the durability of mixtures containing

high percentages of RAP is affected by increasing the volume of virgin binder or decreasing the

performance grade of virgin asphalt. They compared the results of 0, 25 and 50 percent RAP mixes

with PG67-22 and PG 58-28 virgin asphalt binders. Their research finding showed that the fatigue

life of both 25 and 50 percent RAP improved when they used a soft binder grade. They observed

13

that for 25 percent RAP mixes, by increasing effective virgin binder content the number of cycles

to failure also increased but this trend was not seen for the 50 percent RAP binder blends.

Needed research

There has been limited research carried out on the effect of the addition of rejuvenators on the

properties of aged binders in RAP and their influence on the performance of HMA with a high

percentage of RAP content. There is a need to investigate the laboratory properties of HMA with

a high RAP content with rejuvenators, and compare them to the properties of regularly used mixes

with relatively low percentage of RAP. This present study is an attempt in this direction. One

unique feature of this study was the use of non-destructive test for evaluation of recycled mixes

before and after moisture conditioning. The creep compliance test was conducted to evaluate the

cracking potential at low temperature and the SCB test was selected to investigate the fracture

resistance of the mixes with high RAP. The Moisture Induced Stress Tester (MIST) was carried

out to assess moisture susceptibility of the HMA mixes with high RAP. PG grading of asphalt

extracted from recycled mixes was conducted to determine the relevant properties at low,

intermediate and high temperatures.

14

3 EXPERIMENTAL INVESTIGATIONS

Experimental plan

Figure 1 shows the experimental plan adopted for this study. It involves preparation and testing of

150 mm diameter and 38.1mm height samples, determining seismic modulus with the UPV at two

different temperatures of -10°C and 25°C for pre-MIST and post-MIST samples. This was

performed to study the properties of rejuvenated and non-rejuvenated RAP mixes and their

susceptibility towards moisture damage, and also to evaluate their effect on stiffness at low

temperatures. The ITS and SCB test were conducted to study the strength and fracture energy of

the recycled mixes at 25°C.

Figure 1: Experimental plan

Before framing this experimental plan, a plan was proposed in conducting dynamic modulus test

in addition to ultrasonic pulse velocity test. But due to the problem of placing the LVDT (Linear

20% RAP RAP 50% RAP + WVO 50% RAP + SV

(Control)

Preparation of 114.3mm, 150mm

diameter HMA samples with 7±1% Air

voids

Preparation of 114.3mm, 150mm diameter

HMA samples with 4±1% Air voids

Determine Seismic Modulus by

Ultrasonic pulse velocity test (UPV) at

25°C and -10°C.

Determine creep compliance at -10°C

Subject samples to MIST (Moisture

Induced Stress testing) at 25°C, 15000

cycles

Determine Seismic Modulus by

Ultrasonic pulse velocity test (UPV) at

25°C and -10°C.

Extract Binder from loose mix for different

rejuvenator dosages

Conduct Semicircular bend test (SCB)

Determine the Indirect tensile strength

Extract Binder from loose mix

Cut 114.3mm tall samples to two 38.1

mm samples and check volumetric

properties

Cut 114.3mm tall samples to two 50.8 mm

samples check volumetric properties

Cut each 50.8 mm samples to two

semicircular halves and check volumetric

properties

20% RAP+ WVO

WV- Waste vegetable oil

SV- Sylvaroad

MIST – Moisture Induced Stress Tester

15

Variable Differential Transformer) mounts on the specimen, i.e. due to adhesion failure of LVDT

mounts on the specimen, caused by the vacuum grease couplant used in the UPV test, the dynamic

modulus test was not carried out, and the modulus values for the samples were computed from

seismic modulus obtained from the UPV test.

16

4 MATERIALS AND METHODS

Gradation

The HMA mixes were prepared using RAP that was milled from in-service pavements in the city

of West Brook, Maine, USA, and aggregates and asphalt binder provided by Maine DOT. The

gradations adopted for the 20% RAP and 50% RAP mixes are shown in Table 1.The different

stock pile materials used for this study consisted of 12.5mm, 9.5mm, Dry screen stone (DSS), Wet

screen stone (WSS) and Sand and RAP. The percentages of these individual materials used for the

samples are shown in Table 2.

Table 1: Gradation of materials used in the present Study

Sieve Size (mm)

Percent Passing

20% RAP 50% RAP Specification

Limits

19 100 100 100

12.5 99 99 92-100

9.5 87 89 80-90

4.75 59 57 52-66

2.36 45 44 41-49

1.18 34 34 30-38

600 22 23 19-25

300 12 13 10-14

150 6 8 5-9

75 4 6 2-6

Table 2: Individual aggregate proportions and asphalt content

Material Aggregate Proportions

20% RAP 50%RAP

12.5mm 26 20

9.5mm 14 16

WSS 17 0

DSS 5 0

Sand 18 14

RAP 20 50

Virgin asphalt

content (%) 4.7 3

17

Virgin Binder

The virgin binder used for preparing the mix was of grade PG 58-28 (provided by MDOT), which

contained Evotherm (http://www.ingevity.com/markets/asphalt-and-paving/), a warm mix

additive. A lower binder grade was selected by Maine DOT so as to compensate for the aged RAP

binder. From binder extraction by ignition, the average binder content in the RAP mix was found

to be 5.6 %. For 50% RAP mixes, the virgin binder content requirement was 3% of the total HMA

mix [39], assuming the rest of the binder comes from the rejuvenated RAP material. Note that the

NCAT recommendation (NCAT report No. 12-03) [40] of using slightly higher binder percentage

when using higher RAP percentages was used. For 20% RAP mixes, the virgin binder content

requirement was determined to be 4.7 % with 0.9% binder coming from RAP.

Rejuvenators

Rejuvenators are additives which are formulated to restore original properties like relaxation,

ductility, cohesive and adhesive properties of aged (oxidized) asphalt binders, by restoring its

original ratio of asphaltene to maltene. A rejuvenator usually contains high proportion of low

viscosity maltene constituents to help restore the balance between maltene and asphaltenes of RAP

binder that are changed during the aging process.

Waste vegetable oil (WV oil)

Waste vegetable oil, also known as cooking oil is derived from waste food frying oil. It is also

sometimes referred to as “yellow grease”. The product used for this study consisted of peanut,

sunflower, and canola oils, with large concentrations of oleic and linolic acids. It had low free fatty

acids content and less than 2% moisture content in it. The optimum dosage of this rejuvenator was

found to be 10.4 % by percentage weight of estimated asphalt binder from RAP, from penetration

test results as shown in Figure 2. HMA with 20% and 50% recycled mixes for SCB samples were

prepared using 2%, 6%, 10% and 15% rejuvenator dosages. Table 3 represents the SARA

composition for Waste vegetable Oil.

18

Table 3: Saturates, Asphaltenes, Resins, Aromatics (SARA) table for waste vegetable oil

(Frank, 2014 [41])

Rejuvenator Asphaltenes Saturates Resins Aromatics

Yellow Grease - 0.01 32.4 67.5

Figure 2: Optimum rejuvenator dosage for waste vegetable oil

Sylvaroad (SV)

Sylvaroad is a patented chemical additive produced from bio-renewable raw materials and

marketed by Arizona Chemical (http://www.sylvaroad.com/). It is a pine chemical, derived from

co-product of pulp and paper industry. This chemical is non-hazardous, bio-based and renewable.

It has also been found that this rejuvenator is capable of reducing the stiffness of the RAP binder

while using higher RAP percentages in the field (Rotterdam, NL). A rejuvenator dosage of 8% by

weight of the estimated RAP binder was used according to the manufacturer’s recommendation.

WV Oil

0

20

40

60

80

100

120

140

160

0% 5% 10% 15% 20%

Pen

etra

tio

n, d

mm

Dosage, %

Target penetration

19

5 MIX PROCEDURE

The mixing procedure adopted for this study were different for different mix types. The mixing

procedures are explained in the following sections:

20% RAP and 50 % RAP without rejuvenators

The sequence of mixing for the 20% RAP and 50% RAP recycled mixes without rejuvenator

involved drying and heating the virgin aggregates at 1500C for 4 hours prior to mixing. The RAP

was heated at a temperature of 1500C for 1.5 hours. The virgin binder was heated to mixing

temperature (1500C) for 4 hours, with stirring to ensure homogeneous mixing. The RAP was added

to the hot aggregates followed by the addition of binder and mixing for 2 minutes. The loose HMA

mix was then aged at 1500C for 4 hours prior to compaction. The binder content in RAP and the

virgin binder content added to the mixes are shown in Table 4.

Table 4: Virgin and RAP Binder content in HMA Mixes

20% RAP 50% RAP

RAP Binder Content, % (Ignition method) 5.6 5.6

Virgin Binder Content, % 4.7 3

50% RAP with waste vegetable oil rejuvenator

For recycled mixes with waste vegetable oil as rejuvenator, the virgin aggregates and virgin binder

were heated at a temperature of 150°C for 4 hours prior to mixing. The RAP was warmed up to a

temperature of 150°C for 1.5 hours before mixing. The mixing process involved addition of

rejuvenator (kept at ambient temperature) to RAP and then mixing for 30 s. Virgin aggregates

were then added and mixed for 60 s. This was followed by addition of hot virgin binder and mixing

for another 90 s. The mix was aged at 150°C for 4 hours before compaction. Table 5 represents

the Waste Vegetable Oil rejuvenator content calculation. For SCB test, the loose mix were

prepared and aged as mentioned above using four rejuvenator dosages of 2%, 6%, 10% and 15%.

The loose mixes with four rejuvenator contents were subjected to an additional long term aging at

20

135°C for 24 hours before compaction as they appeared to give more realistic results compared to

the AASHTO R30 long-term aging procedure [42].

Table 5: Waste Vegetable Oil rejuvenator content calculation

RAP Binder Content,% (Ignition Method) 5.6

Virgin Binder Content, %

50% of Binder content available in RAP +

0.2% (NCAT Report No.12-05)[39]

50% of 5.6% + 0.2% = 3

Rejuvenator content (% of RAP binder) 10.4

RAP with rejuvenator Sylvaroad (SV)

For recycled mixes with Sylvaroad as rejuvenator, the virgin aggregates were heated at a

temperature of 1800C (to compensate for the lower RAP temperature) for 8 hours prior to mixing.

The RAP was warmed up at a temperature of 1200C [110-1300C] for 2.5 hours and the virgin

binder was heated at the mixing temperature (1500C) for 3 hours and stirred to have homogeneous

mix before mixing. Sylvaroad rejuvenator (kept at ambient temperature) at a dosage of 8% by mass

of RAP binder was added to the RAP and mixed for 30 seconds. The virgin aggregates were then

added and mixed for 60 seconds. Hot virgin binder was then added and mixed for another 90

seconds. This loose mix was aged at 150°C for 4 hours before compaction. Table 6 shows the

Sylvaroad rejuvenator content calculation

Table 6: Sylvaroad Rejuvenator Content Calculation

RAP Binder Content, % (Ignition Method) 5.6 %

Virgin Binder Content, %

50% of Binder content available in RAP +

0.2% (NCAT Report No.12-05)

=50% of 5.6% + 0.2% = 3%

Rejuvenator content (% of RAP binder) 8% of RAP binder

21

6 COMPACTION

The compaction of the 20% and 50 % RAP mixes were carried out using a Superpave gyratory

(ASTM D 4013). Figure 3a shows the Superpave gyratory compactor used for this study. The loose

mixes after the required period, were compacted with target air void content of 7±1%, for non-

SCB samples and 4±1% for SCB samples. The gyratory compactor was programmed such that the

loose mix would be compacted with preset target height (114.3mm) or 100 gyrations, whichever

is achieved first. The compacted samples had a diameter of 150mm and a height of 114.3mm. The

compacted samples were then cut to 38.1mm thick samples for non-SCB testing and 50 mm

thickness for SCB testing using saw cutter as shown in Figure 3b. The volumetric properties such

as Gmb and Gmm were determined (ASTM D2041 and ASTM D2726) and air void contents of the

each cut samples were calculated.

Figure 3: (a) Superpave Gyratory compactor and (b) Saw cutter

a) SUPERPAVE GYRATORY

COMPACTOR

b) SAW CUTTER

MOLD SAMPLE

22

7 TEST METHODS

Ultrasonic pulse velocity

Ultrasonic Pulse Velocity test is a method of non-destructive evaluation of a HMA specimen based

on wave propagation technique. In this test method, a piezoelectric crystal is used, which converts

electrical energy to an ultrasonic shock wave. This shock wave or ultrasound is then transmitted

from the transducer through the specimen and is then collected at the receiving transducer, which

converts this shock wave back to an electrical pulse. The velocity of this pulse through a medium

is dependent upon the material properties of the specimen such as its elastic properties and density

of the medium. The test was conducted according to ASTM C597-09. For this study, six specimens

were prepared for each mix type (24 specimens in total) of 150 mm diameter and 38 mm thickness.

The equipment used in the study is a V-meter MK II, made by James Instruments, Inc, which

generate ultrasonic pulse waves at a frequency of 54 KHz. The transducers were placed in direct

transmission position for transmission and reception of these pulse waves. Damping pads and

vacuum grease were used as couplant to ensure full contact between the transducer and specimen.

A loading plate was placed on top of the transducer to ensure uniform pressure on the specimen.

For calibration purposes, a specimen of known dimension was used. Figure 4 shows the apparatus

used and experimental setup for the UPV test. The design modulus was obtained from seismic

modulus (Nazarian et al., 2002) [43].

The specimen dimension was determined for each sample, and the compression wave (P-wave)

velocity, Vp was then calculated from this equation:

Vp= H* tv (1)

Where H is the height of the specimen and tv is the corresponding travel time (mean of four

transmission time readings per sample). The constraint modulus, MV, was then calculated using

Mv = ρ*Vp (2)

Where ρ is the bulk density of the specimen in g/cc.

The constraint modulus was then converted to Young’s modulus, EV through a theoretically

corrected relationship in the form of

23

𝐸𝑣=𝑀𝑣∗((1+μ)∗(1−2μ))/((1−𝜇)) (3)

Where Evis young’s modulus and μ is Poisson’s ratio. The Poisson’s ratio for all mixes was

considered as 0.35.

In this study, the UPV test was conducted at two temperatures. 25°C and-10°C. These temperatures

were selected to study the performance of the mix under expected fatigue cracking and low

temperature cracking conditions respectively.

Figure 4: Ultrasonic device

Creep compliance and Indirect Tensile Strength test (ITS)

Creep compliance is a test method for characterizing the stiffness of material. It is used to

determine the cracking potential of the asphalt mixes at low temperatures. For this study, the test

was conducted according to AASHTO T322 and Indirect Tensile strength was performed

according to ASTM D6931. These tests were used to evaluate the low temperature cracking

APPARATUS EXPERIMENTAL SETUP

Transmitter

Receiver

24

potential and the strength of the recycled mixes. A total of three specimens for each mix type were

used for each test. Figure 5 shows the experimental set up for creep compliance.

Creep compliance is defined as a ratio of time-dependent strain to the applied stress. It is

determined by applying a static compression load to cause sufficient horizontal deformation in the

sample between 0.00125 mm and 0.0190 mm (linear viscoelastic range). For a specimen with a

gauge length of 38 mm, the corresponding strain range is recommended to be between 33x10-6 and

500 x10-6 mm/mm (viscoelastic range) (AASHTO T322). The test is run for a period of 1,000

seconds at a temperature of -10°C, after a conditioning period of 3 hours. During the test, the

horizontal and vertical deformations are measured on each face of the specimen using LVDT. For

cases in which the strain on one face was higher (> allowable) than that in the other, the average

strain value was chosen based on whichever face had a strain value within the viscoelastic range

(33-500 µm).

Creep compliance was calculated by using the following equation:

𝐷(𝑡) = 1 +

∆𝑋𝑡 × 𝑑 × 𝑏

𝑃 × 𝐺𝐿× 𝐶𝑐𝑚𝑝𝑙 (4)

Where,

ΔXt is the mean horizontal deflection (mm)

d is the diameter of the sample (mm)

b is the thickness of the sample (mm)

P is the creep load (kN)

GL is the gage length over which deformation is measured (mm)

Ccmpl is the compliance factor

C𝑐𝑚𝑝𝑙 = 0.6354 (

𝑋

𝑌)

−1

− 0.332 (5)

Where X/Y is the ratio of horizontal to vertical deflection

25

The limits of Ccmpl are as follows:

[0.704 − 0.213 (

𝑏

𝑑)] ≤ 𝐶𝑐𝑚𝑝𝑙 ≤ [1.566 − 0.195 (

𝑏

𝑑)] (6)

Figure 5: Creep compliance test setup

Indirect Tensile Strength (ITS)

The Indirect tensile test (ASTM D6931) was used to determine the strength of the asphalt mixes.

The test was conducted by loading a cylindrical across its vertical diametric plane at a specified

rate of deformation (50 mm per minute) and at a test temperature of 25°C. The peak load at failure

was recorded and was used to calculate the ITS strength of the specimen. The Indirect Tensile

strength of the specimen is given by equation 7:

𝐼𝑛𝑑𝑖𝑟𝑒𝑐𝑡 𝑇𝑒𝑛𝑠𝑖𝑙𝑒 𝑆𝑡𝑟𝑒𝑛𝑔𝑡ℎ (𝐼𝑇𝑆) =

2𝑃

𝜋𝑑𝑡 (7)

Where,

ITS = Indirect Tensile Strength (kPa)

P = Peak Load applied under failure (N)

D = Diameter of the specimen (mm) and

t =Thickness of the specimen (mm)

26

Moisture Induced Stress Tester (MIST)

The Moisture Induced Stress Test (MIST) (equipment shown in Figure 6) simulates the generation

of pore water pressure which is generated in a saturated pavement under repeated traffic loads. In

this test, water is forced in and out of the samples by applying compressed air through a bladder

assembly. The samples for this study were subjected to 15,000 cycles of loading with water

pressure at 30 psi pressure and 25℃ test temperature. The above conditions were selected over

an ASTM 7870 protocol of 60°C at 3,500 cycles and 40 psi to ensure the integrity of the specimen

and the ability to test them after conditioning with the MIST.

Figure 6: Moisture Induced Stress Test equipment

Semi Circular Bend test (SCB)

The Semi Circular Bend Test (SCB) was used to determine the fracture energy and fracture

toughness of the asphalt mixtures at intermediate temperature. The SCB test was performed

according to the AASHTO standard (Al-Qadi et al., 2015) [44] at 25oC. For this test, a semicircular

disc of HMA, 150 mm in diameter and 25 mm thick, was tested in a 3-point bending mode as

shown in Figure 7. Before conducting the test, a notch was created at the center for all samples,

27

for a depth of 15 mm from the flat face of the specimen to initiate the crack propagation. The test

was performed by imposing a small contact load of 0.1 ± 0.01 kN and then by loading at a rate of

50mm/min. The test was stopped once the load dropped below 0.1 kN.

The total work of fracture Wf was calculated by dividing the load-displacement data into two parts

i.e. curve prior to peak load and the curve after the peak load and then numerically integrating the

total area under the two parts.

The total work of fracture is calculated using the integral equation

W𝑓 = ∫ 𝑃1(𝑢)𝑑𝑢 + ∫ 𝑃2(𝑢)𝑑𝑢

𝑢𝑓𝑖𝑛𝑎𝑙

𝑢0

𝑢0

0

(5)

Where Ufinal is the displacement at 0.1kN cut-off load

U0 is the displacement at peak load (kN)

The fracture energy Gf was then found by dividing the work of fracture by the ligament area of

the SCB specimen prior to testing

G𝑓 =

𝑊𝑓

𝐴𝑟𝑒𝑎𝑙𝑖𝑔× 106 (6)

Where:

Gf = fracture energy (Joules/m2)

Wf = work of fracture (Joules)

P =load (kN)

Area lig = ligament length x×t

t = specimen thickness (mm)

The Flexibility Index (FI) is calculated from the parameters obtained from the load displacement

curve.

𝐹𝐼 =𝐺𝑓

|𝑚|× 𝐴

28

Where

FI= Flexibility Index

|m|= absolute value of post-peak load slope m (kN/mm)

A= 0.01

Figure 7: Semicircular bend test (SCB) and specimen

Asphalt binder extraction

The binder from 20% RAP and 50 % RAP - rejuvenated and non-rejuvenated loose mix samples

were extracted using solvent extraction procedure (ASTM D5405). For this method, toluene

solution was used as a solvent to extract asphalt binder from the aggregates of the loose recycled

mixes. This extracted asphalt with the toluene was transferred into a rotating distillation flask of

the rotary evaporator, which was partially immersed in a heated oil bath. This solution was

subjected to partial vacuum and a flow of nitrogen gas to separate asphalt binder from toluene. The

recovered asphalt was then subjected to penetration test, Rolling Thin Film Oven (RTFO) test and

Direct Shear Rheometer (DSR) test to find the penetration grade, high, low and intermediate

temperatures of the extracted binder.

29

8 HIRSCH MODEL

The Hirsch model is a semi-empirical method of predicting modulus of HMA. Christensen et al.

(2003) developed the application of the Hirsch model to HMA mixes. They had examined four

different models based on the law of mixtures, parallel model and selected the model that

incorporates the binder modulus, VMA, and VFA because it provides accurate results in the

simplest form in the prediction of the modulus of HMA. For this study the Hirsch model was used

to back calculate the shear complex modulus of the binder (G*), by using the design modulus

values obtained from ultrasonic pulse velocity test and the volumetric properties by assuming that

the bulk specific gravity of the aggregate (Gsb) is equal to effective specific gravity of the aggregate

(Gse). The following equations were used to back calculate the complex modulus of the binder

(G*) from Hirsch model.

E* = 𝑃𝑐 × [4200000 (1 −𝑉𝑀𝐴

100) + 3 |𝐺∗|𝑏𝑖𝑛𝑑𝑒𝑟 (

𝑉𝐹𝐴∗𝑉𝑀𝐴

10000) +

1 − 𝑃𝑐

[(1 −

𝑉𝑀𝐴100 )

4200000 +𝑉𝑀𝐴

3 |𝐺∗|𝑏𝑖𝑛𝑑𝑒𝑟 𝑉𝐹𝐴]

(7)

𝑃𝑐 =

(20 +3 |𝐺∗| 𝑉𝐹𝐴

𝑉𝑀𝐴 )

0.58

650 + (3 |𝐺∗| 𝑉𝐹𝐴

𝑉𝑀𝐴 )

0.58 (8)

Where

E*= Modulus (MPa) (Derived from Seismic Modulus [37])

G*= Shear Complex Modulus, (kPa)

VMA= Voids in Mineral Aggregates, %

VFA= Voids Filled with Asphalt, %

30

9 RESULTS AND ANALYSIS

A total of six samples were made for each mix type of 20% RAP (20R), 50% RAP (50R), 50%

RAP with waste vegetable oil (50R-WV) and 50% RAP with Sylvaroad (50R-SV), of which three

samples were used for ultrasonic pulse velocity test and the other three for creep compliance test.

The air voids of the samples are shown in Table 7. Analysis of Variance (ANOVA), a statistical

test used to determine the equality between the means of several groups, was carried out to

determine if there is significant difference among the test properties of the different mix types

using Stat tools software [44].

Table 7: Air Void Distribution among samples

Specimen

No

Air Voids, %

20 R 50 R 50R-WV 50 R-SV

1 7.1 6.7 6.9 6.6

2 7.2 6.8 6.3 6.8

3 7.3 7 7.1 6.9

4 7.5 7.1 6.3 7.6

5 7.6 7.4 6.8 7.7

6 8 6 6.9 6

The following designation were used to denote the Mix type in this project: 20R - 20% RAP mix,

50R - 50% RAP mix, 50R-WV - 50% RAP mix with Waste Vegetable oil Rejuvenator 50R- SV -

50% RAP mix with Sylvaroad rejuvenator, 100R- 100% RAP.

Ultrasonic Pulse Velocity (UPV) test result

The ultrasonic pulse velocity test was performed at two temperatures to study the effects of

rejuvenated blends at low temperatures and intermediate temperatures. Seismic and the design

moduli values obtained for the different mixes are shown in Table 8. The UPV test results reveal

that the moduli values are higher at -10°C. Figure 8 shows the average Design moduli values at

both 25°C and -10°C. Figure 9 shows the change in moduli values when the temperature was

changed from 25°C to -10°C for the different mixes. While the trends in changes in moduli values

with temperatures are similar, the plots do show that the rejuvenators are effective in lowering the

stiffness of the 50R mix, and hence in rejuvenating the recycled mix. The results were analyzed to

determine whether a significant difference between the different mixes exists. Table 8 shows that

the 50% RAP mixes are ranked A (indicating a higher modulus) followed by the rejuvenator mixes

which are ranked B, and the 20% RAP mixes, which are ranked as C.

31

Table 8: Seismic and Design Moduli at 25°C and -10°C

(Note: A higher rank indicates a higher Modulus)

Specimen

ID

Avg.

Travel

Time at 25

°C

(microsec)

Avg.

Travel

Time at -

10 °C

(microsec)

Seismic

Modulus

at 25 °C

(MPa)

Seismic

Modulus

at -10 °C

(MPa)

Design

Modulus

at 25 °C

(MPa)

Design

Modulus

at -10 °C

(MPa)

RANKI-

NG

ID t t SM SM DM DM

20R-1 11.3 9.9 12036 15831 3761 4947

C

20R-2 11.6 9.9 11807 16152 3690 5048

20R-3 11.4 9.8 12042 16284 3763 5089

20R-4 11.4 9.8 12226 16483 3821 5151

20R-5 11.7 9.9 11327 15808 3540 4940

20R-6 11.6 9.8 11620 16519 3631 5162

50R-1 10.8 9.4 13784 18221 4308 5694

A

50R-2 10.6 9.3 14416 18652 4505 5829

50R-3 10.6 9.3 13996 18269 4374 5709

50R-4 10.7 9.3 13950 18380 4359 5744

50R-5 10.6 9.4 14117 18144 4412 5670

50R-6 10.6 9.3 14019 18310 4381 5722

50R-WV-1 11.0 9.5 13071 17792 4085 5560

B

50R-WV-2 10.9 9.3 13582 18388 4244 5746

50R-WV-3 11.1 9.5 13047 17839 4077 5575

50R-WV-4 11.0 9.4 13298 18391 4156 5747

50R-WV-5 11.2 9.5 12740 17564 3981 5489

50R-WV-6 11.0 9.4 12818 18020 4006 5631

50R-SV-1 11.1 9.4 12937 18069 4043 5647

B

50R-SV-2 11.0 9.3 12979 18157 4056 5674

50R-SV-3 11.0 9.3 13188 18268 4121 5709

50R-SV-4 11.0 9.3 13184 18345 4120 5733

50R-SV-5 11.0 9.4 13124 18054 4101 5642

50R-SV-6 11.0 9.4 13177 18031 4118 5635

32

Figure 8: Average Design Modulus: A) 25°C; B) -10°C

0

1000

2000

3000

4000

5000

6000

7000

20 R 50 R 50 R-WV 50 R-SV

Des

ign M

odulu

s, M

Pa

Mix type

A) Average Design Modulus (MPa) at 25°C

0

1000

2000

3000

4000

5000

6000

7000

20 R 50 R 50 R-WV 50 R-SV

Des

ign M

odulu

s, M

Pa

Mix type

B) Average design Modulus (MPa) at -10°C

33

Figure 9: Plots of temperature versus modulus for the different mixes

Table 9: ANOVA for Design Modulus (DM) at 25°C:

Where 50RA - 50% RAP Waste Vegetable Oil rejuvenator; 50RB - 50% RAP Sylvaroad

rejuvenator.

0

1000

2000

3000

4000

5000

6000

7000

20R 50R 50R-WV 50R-SV

Des

ign M

od

ulu

s, M

Pa

Mix type

Temperature vs Design Modulus

25°C -10°C

34

Table 10: ANOVA for Design Modulus (DM) at -10°C

Where 50RA - 50% RAP Waste Vegetable Oil rejuvenator; 50RB - 50% RAP Sylvaroad

rejuvenator.

Table 9 and 10 confirms that the results are significantly different. Table 9 also shows that while

there are differences between most of the mixes, there is no significant difference between the

moduli of the 50R-WV and 50R-SV. Table 10 shows that at -10°C there is no significant difference

between the moduli of 50R and 50R-SV and also 50R and 50R-WV

MIST test results

Three samples were chosen for each mix type with similar air void (7±1) content. These samples

were then subjected to MIST. Table 11 shows the results of the UPV tests conducted before and

after MIST test. An increase in the moduli values were observed after MIST test. Figure 10 shows

the pre and post MIST modulus values, as well as the decrease in air voids for each of the samples

of the different mixes. It is noted that even though the change in voids are the same or lower than

the control and the 50R mixes, the rejuvenated mix samples have a comparatively higher change

in modulus after the MIST process (that is, higher difference between pre and post MIST modulus).

These could be due to either or both of two reasons. First, the presence of the rejuvenator could be

facilitating the compaction of the mixes in such a way that the aggregate structure is improved,

which leads to an increase in the modulus. The second possibility is that the action of the MIST

conditioning process is reversing the effect of rejuvenation, and making the mix similar to the 50R

mix. This possibility is being raised because it is noted that the post MIST modulus of the

35

rejuvenated mix samples are almost at the same level as that of the post MIST modulus of the 50R

samples. Table 12 shows the bulk specific gravity (BSG) and Air void (%) results before and after

MIST test. We can see that there is an increase in density of the mix after MIST test. It can be seen

from Table 11 that the difference between the pre and post-MIST samples were similar for the

rejuvenated mixes, and higher than that of the other mixes. The 20R samples had the lowest

difference between the pre and the post MIST samples. Further study is required to evaluate the

change in modulus in rejuvenated samples due to the MIST process.

Table 11: Moisture Induced Sensitivity Tester (MIST) results on Moduli Values of RAP

mixes

(Note: A higher rank indicates a higher moisture effect)

Sample ID

Design Modulus,

before moisture

effect, MPa

Design Modulus

after moisture

effect, MPa

Difference

between Pre and

Post Moisture

Test, %

( 𝑃𝑜𝑠𝑡 − 𝑃𝑟𝑒

𝑃𝑟𝑒)%

Ranking

20R-2 3690 3956 7.21

B 20R-3 3778 3914 3.59

20R-6 3631 3888 7.08

50R-1 4291 4344 1.23

C 50R-3 4378 4407 0.65

50R-5 4333 4374 0.95

50R-WV-1 4085 4563 11.70

A 50R-WV-5 3981 4531 13.81

50R-WV-6 4006 4579 14.32

50R-SV-1 4043 4496 11.21

A 50R-SV-2 4056 4491 10.73

50R-SV-5 4101 4596 12.07

36

Table 12: Density comparison (PreMIST vs PostMIST)

Sample ID

Pre-MIST Post-MIST BSG

Difference

(Pre-Post) % BSG AV (%) BSG AV (%)

20R-2 2.293 7.08 2.312 6.3 0.8

20R-3 2.291 7.2 2.307 6.5 0.7

20R-6 2.287 7.3 2.300 6.8 0.5

50R-1 2.307 7.0 2.324 6.3 0.7

50R-3 2.305 7.1 2.318 6.6 0.5

50R-5 2.299 7.4 2.321 6.4 1.0

50R-WV-1 2.307 6.9 2.322 6.3 0.7

50R-WV-5 2.309 6.8 2.324 6.2 0.6

50R-WV-6 2.323 6.3 2.332 5.9 0.4

50R-SV-1 2.296 7.6 2.297 7.5 0.0

50R-SV-2 2.293 7.7 2.299 7.4 0.3

50R-SV-5 2.312 6.9 2.326 6.4 0.6

Figure 10: Comparison of Pre, Post MIST design Moduli and Air voids

37

Table 13: ANOVA for Pre-MIST DM at 25C

Table 14: ANOVA for Post-MIST DM at 25C

From Table 13 and 14 we can see that the all the mixes are significantly different except the

rejuvenator mixes. There is no significant difference observed between the 50R-SV samples and

50R-WV samples. This also tells us that both rejuvenated mixes are affected to the same extent by

the MIST process.

38

Penetration test results

The extracted binder from the loose mix samples were tested for penetration (ASTM D5-06).

Penetration test was performed to evaluate the effects of high percentages of RAP binder

containing rejuvenators on the properties of virgin binder and overall binder properties. The results

are presented in Table 15. The observed penetration values for the 100% RAP and the 50%RAP

are low as expected. This indicates that 100 % RAP binder and 50 % RAP binder are stiffer

compared to the 50% RAP rejuvenated binders and 20% RAP binders. The penetration results

reveal that the rejuvenators are effective in reducing the overall stiffness of the aged RAP binders.

Of the two rejuvenators, the Sylvaroad rejuvenator was found to be more effective in reducing the

overall stiffness of the binder and hence resulted in a higher penetration values compared to the

waste vegetable oil. Figure 11 represents the penetration values obtained for the different mix type.

The results of Mean Separation test are also shown in Table 15. It can be seen from table that the

50R-SV mix is ranked A (which indicates more soft mix than others), the 50R-WV is ranked B,

followed by the control mix (20R) which is ranked as C and 50% RAP and 100% RAP which are

ranked as D.

Table 15: Penetration Test Results on Extracted Binder from HMA mixes

(Note: A higher rank indicates a higher penetration value)

Extracted

Asphalt

Temperature Penetration 1/10mm

Ranking °C

Measurement

1

Measurement

2

Measurement

3 Mean

100% RAP 25 32 32 25 30 D

20% RAP 25 50 51 41 47 C

50% RAP 25 25 31 27 28 D

50%RAP+WVO 25 62 65 66 64 B

50%RAP+ SV 25 80 86 81 82 A

39

Figure 11: Penetration results

Table 16: ANOVA for Penetration at 25°C

From Table 16 we can say that the mixes are significantly different except that we observe no

significant difference between penetration values of 100% RAP and 50% RAP binders.

0

10

20

30

40

50

60

70

80

90

100

100R 20R 50R 50R+WV 50RR+ SV

Pen

etra

tio

n (

1/1

0)m

m

Mix type

Penetration results at 25°C

40

Performance Grade (PG) results:

The binder was extracted from the different mixes and tested for Performance Grade (PG). The

results are tabulated in Table 17. It can be seen from the table, that 50% RAP binder and 100%

RAP binder are severely aged and graded as PG 81-24 and PG 78-24 respectively, but with the

addition of Waste vegetable oil and Sylvaroad rejuvenators, the binders were graded as PG-68-31

and PG 66-34 respectively. This indicates that the addition of these rejuvenators are capable of

reducing the stiffness of the aged binder and probably restoring the original properties of RAP

binder. From Figure 12, it can be inferred that of the two rejuvenators, Sylvaroad at the content

used in this study, is more effective in reducing the stiffness of the aged binder compared to waste

vegetable oil. The binders of the rejuvenated blend shows better characteristics (higher PG sum)

in comparison to that of the control mix.

Table 17: Performance Grade (PG) -Low and High results

Binder State Continuous high PG, ℃ Continuous low PG, ℃ PG sum,℃

Virgin binder 58.0 -28.0 86.0

20R 71.8 -27.1 98.9

50R-WV 68.1 -31.3 99.4

50R-SV 66.4 -34.0 100.4

50R 81.1 -24.0 105.0

RAP 78.2 -24.1 102.3

Figure 12: Continuous PG- Low High

78.258.0

71.881.1

68.1 66.4

-24.1 -28.0 -27.1 -24.0 -31.3 -34.0

-40.0

-20.0

0.0

20.0

40.0

60.0

80.0

100.0

RAP PG58-28 20R+PG58-28 50R+PG58-28 50R-WV+PG58-28 50R-SV+PG58-28

Per

form

an

ce G

rad

e (P

G),

°C

Continous PG- low high chart

High Low

41

Back calculated Hirsch model results

Hirsch model was used to back calculate the complex modulus (G*) from the design modulus

values obtained from the Ultrasonic pulse velocity (UPV) test. The volumetric calculations were

made to calculate the Voids in Mineral Aggregates (VMA) and Voids Filled with Asphalt (VFA)

of the mix. For this calculation of volumetric properties, it was assumed that the bulk specific

gravity (Gsb) and the effective specific gravity (Gse) are same. This assumption was made to bypass

the problem of finding the Gsb of coated RAP particles. The volumetric properties of the different

mixes are shown in Table 18. The back calculated values are tabulated in Table 19. From the table,

it can be observed that there is difference in moduli values at 25 C and -10C between the different

mixes. It can also be observed that there is significant difference in moduli value at two different

temperatures for each of the rejuvenated mixes in comparison with the 50% RAP and 100% RAP

mixes. This proves the significant influence of the rejuvenator on the moduli values. Table 20

shows the ANOVA results obtained from the G* data. It can be seen from Table 21 that the 50R

mix is ranked as A, 50R-SV mix is ranked B followed by 50% RAP- WV, ranked C and finally

20% RAP which is ranked D. From Table 21 we can say that the mixes have significantly different

back calculated complex modulus (G*)

Table 18: volumetric properties

42

Table 19: Back calculated G* using Hirsch Model

Mix ID Back calculated G*,

kPa at 25°C

Back calculated G*,

kPa at -10°C

20R-2 2,732 5,308

20R-3 2,908 5,479

20R-6 2,722 5,745

50R-1 3,783 6,852

50R-3 3,988 7,059

50R-5 4,017 7,081

50R-WV-1 3,362 6,568

50R-WV-5 3,155 6,316

50R-WV-6 2,994 6,258

50R-SV-1 3,572 7,388

50R-SV-2 3,648 7,581

50R-SV-5 3,405 6,814

Table 20: ANOVA for back calculated G* at 25C

43

Table 21: Ranking of Mix types

(Note: A higher rank indicates a higher complex moduli)

Mix ID Backcalculated G*, kPa

from E* AT 25C

Ranking (High to

Low)

20%-RAP

2,732

D 2,908

2,722

50%-RAP

3,783

A

3,987

4,016

50%-RAP-WV

3,362

C

3,155

2,994

50%-RAP-SV

3,572

B

3,648

3,405

Indirect Tensile Strength (ITS) results

The ITS test results were obtained for samples tested at 25°C. From Table 22, it can be observed

that the 50% RAP mixes show higher indirect tensile strength values compared to the other mixes.

Table 22: Indirect Tensile Strength Test (ITS) Results

S.No SAMPLE ID Peak Load

(N)

Average

specimen

height(t) (mm)

Average

dia of

specimen

(D) (mm)

ITS (kPa)

1 20R-2 6,130 38.9 152.4 659

2 20R-3 6,512 38.6 152.4 705

3 20R-6 6,868 38.6 152.4 743

- Average 6,503 38.7 152.4 702

4 50R-1 8,469 38.9 152.4 910

5 50R-3 8,140 38.6 152.4 881

6 50R-5 8,928 38.6 152.4 966

- Average 8,512 38.7 152.4 919

7 50R-WV-1 5,765 38.8 152.4 620

8 50R-WV-5 4,893 38.7 152.4 529

9 50R-WV-6 5,445 38.6 152.4 589

- Average 5,368 38.7 152.4 579

10 50R-SV-1 5,222 38.9 152.4 561

11 50R-SV-2 4,875 38.6 152.4 527

12 50R-SV-5 5,792 38.6 152.4 627

- Average 5,296 38.7 152.4 572

44

The rejuvenated mixes on the other hand show lower ITS values than the control and 50% RAP

mixes. This indicates that rejuvenators are effective in reducing the overall stiffness of the mix at

intermediate temperature which is relevant for evaluation of the potential of fatigue cracking.

Figure 14 shows the Indirect Tensile Strength test results for the RAP samples with and without

the addition of rejuvenators. Table 24 shows the Mean separation test results for the ITS data.

From this table it can be seen that 50% RAP has been ranked A, followed by 20% RAP ranked as

B and the rejuvenated mixes ranked as C.

Figure 13: Indirect Tensile Strength (ITS) results

0.0

0.2

0.4

0.6

0.8

1.0

1.2

20R 50R 50R+WV 50R+SV

ITS

(KP

a)

SAMPLE TYPE

Indirect Tensile Strength (ITS) results at 25°C

45

Table 23: ANOVA for ITS at 25 C

Table 23 shows the ANOVA test results for different mix types. We can see from the table that

there is significant difference between all mixes except between 50R-WV and 50R-SV mixes.

Table 24: Ranking for ITS at 25°C

(Note: A higher rank indicates a higher tensile strength)

Mix ID POST MIST Indirect Tensile Strength

(ITS), KPa Ranking

20R 659

B 705

743

50R 910

A 881

966

50R-WV 620

C 529

589

50R-SV 561

C 527

627

Creep compliance results

Creep compliance test is a way to characterize the stiffness of material. A higher stiffness or a low

compliance at low temperature is indicative of a mix with potential of thermal cracking. Table 25

46

shows the results of the creep compliance test. The creep compliance results reveal that the 50%

RAP rejuvenated mixes have better resistance to failure at low temperature in comparison to the

20% RAP and 50% RAP mixes. It can also be seen from Table 25 that the 50% RAP mixes show

low creep compliance values which indicates that this mix has poor resistance to cracking at low

temperature. Statistical analysis could not be done here as we require results of three specimens

that are analyzed simultaneously to reduce variability in determining Poisson’s ratio and creep

compliance. Figure 14 shows the creep compliance value at -10°C.

Table 25: Creep compliance results

Mix type Creep compliance (1/GPa)

20R 0.199

50R 0.113

50R-WV 0.223

50R-SV 0.248

Figure 14: Creep compliance at -10C

0

0.05

0.1

0.15

0.2

0.25

0.3

20R 50R 50R-WV 50R-SV

Cre

ep c

om

pli

ance

(1/G

Pa)

Mix type

Creep compliance (1/GPa) at -10°C

47

Semi Circular Bend (SCB) test results

The SCB test results were obtained for samples with 4-5% air voids and at intermediate test

temperature (25°C). Table 26 and Table 27 show the fracture energy results obtained for 20% RAP

and 50% RAP samples with different waste vegetable oil rejuvenator dosages. From Table 26 we

can infer from the mean values that addition of different rejuvenator dosages do not have

significant influence over 20% RAP mix, whereas from Table 27 we can see that the Fracture

energy drops for 50% RAP as the rejuvenator dosage increases. This indicates the effectiveness of

the waste vegetable oil rejuvenator in softening the mix and thereby reducing the overall stiffness

of 50% RAP mixes. Figure 15 shows the Fracture energy at different rejuvenator dosages for 20%

RAP and 50% RAP. The flexibility index (FI) is a parameter that can indicate the cracking

potential of asphalt concrete mix. A higher FI is desirable for greater resistance against cracking.

We can see from Table 26 that the average flexibility index for 20% RAP remains fairly same for

the different rejuvenator dosages except for 15% rejuvenator dosage where we can see it increases

to 3. This also shows that at 15% rejuvenator dosage, the mix is less brittle in nature with respect

to other dosages. Table 27 shows that for 50% RAP mixes, with the increase in rejuvenator dosage

the flexibility index also increases specifically at 15% dosage. Overall a comparison of 20R and

50R mixes show a higher FI for the 20R (lower RAP content) mixes except at 15% rejuvenator

dosage. This would indicate that the rejuvenators are capable of reducing the stiffness of the 50%

RAP and thereby making it less brittle and more resistance to fatigue cracking. Figures 16 and 17

show the side by side comparison of 20% RAP and 50% RAP fracture energy and FI at different

rejuvenator dosages.

48

Table 26: Fracture Energy test results of 20% RAP

Sample ID Fracture Energy ( Joules/m2)

2% -WV FI 6% -WV FI 10% -WV FI 15% -WV FI

20R-1-1 1854 1.9 2019 3.2 1941 2.5 1898 2.2

20R-1-1A 1849 2.5 2686 3.9 2111 2.1 1841 2.5

20R-1-2 1984 3.9 2869 3.7 1521 1.1 1959 4.9

20R-1-2A 1545 1.5 2277 3.5 2047 1.4 1959 2.3

20R-2-1 1867 1.4 2407 1.8 1640 0.5 2254 2.1

20R-2-1A 1557 1.2 1585 0.8 1387 1.1 2466 2.4

20R-2-2 2000 2.7 2306 1.3 1609 1.4 2185 2.3

20R-2-2A 1827 0.7 2234 1.6 2028 1.8 2057 3.2

20R-3-1 2150 0.8 1634 0.8 2290 3.4 2997 4.3

20R-3-1A 1819 1.3 2478 1.1 2203 2.8 2318 3.9

20R-3-2 1823 1.1 1993 1.2 2903 5 2287 3.7

20R-3-2A 3371 6.1 1570 1.8 2036 1.2 1380 2

Average 1971 2.1 2172 2.1 1976 2 2133 3

SD 471.92 1.56 424.67 1.18 409.21 1.24 393.92 0.99

RANKING A A A A

Table 27: Fracture energy test results for 50% RAP

Sample ID Fracture Energy ( Joules/m2)

2% -WV FI 6% -WV FI 10% -WV FI 15% -WV FI

50R-1-1 1859 0.9 2289 1.5 2177 2.7 1796 4.2

50R-1-1A 2884 1.2 2068 2.5 1953 1.7 2978 5.4

50R-1-2 3600 1.0 2408 2.7 2006 1.4 1483 2.7

50R-1-2A 3164 4.0 2514 1.3 1930 1.5 1334 2.3

50R-2-1 1528 1.0 2142 1.2 1376 1.0 1772 2.9

50R-2-1A 2451 1.6 1709 1.2 1744 0.9 1613 1.9

50R-2-2 1745 0.8 2316 0.9 1719 1.0 1520 1.7

50R-2-2A 1815 1.1 1999 0.9 1565 1.1 1530 2.4

50R-3-1 2074 1.5 1630 1.0 2159 2.7 2344 4.0

50R-3-1A 2236 1.6 2939 1.5 1305 1.7 1855 5.4

50R-3-2 2862 2.1 2315 1.8 2661 2.2 1687 2.8

50R-3-2A 2036 1.5 1371 2.2 2158 1.2 2904 4.6

Average 2354 1.5 2142 1.6 1896 1.6 1901 3.4

SD 642.38 1.03 426.03 0.61 381.65 0.62 547.71 1.31

RANKING A A A A

49

Figure 15: Fracture Energy at different rejuvenator dose for A-20% RAP; 50% RAP.

1700

1800

1900

2000

2100

2200

2300

20R-2 WV 20R-6 WV 20R-10 WV 20R-15 WV

Frac

ture

En

ergy

(J/

m2 )

Sample Type

A - 20% RAP- Fracture Energy (Joules/m2)

1700

1800

1900

2000

2100

2200

2300

2400

2500

50R-2 50R-6 50R-10 50R-15

Frac

ture

En

ergy

(J/

m2 )

Sample Type

B - 50% RAP- Fracture Energy (Joules/m2)

50

Figure 16: Comparison of Fracture Energy of 20% RAP and 50% RAP mixes

Figure 17: Comparison of Flexibility Index of 20% RAP and 50% RAP mixes

1700

1800

1900

2000

2100

2200

2300

2400

2500

2600

2% 6% 10% 15%

Frac

ture

En

ergy

(J/

m2 )

WV-Rejuvenator Dosage

Comparison of 20R and 50R Fracture Energy

20R 50R*20R - 20% RAP with rejuvenator

50R -50% RAP with rejuvenator

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

2% 6% 10% 15%

Flex

ibili

ty In

dex

WV- Rejuvenator Dosage

Comparison of 20R and 50R Flexibility Index (FI)

20R 50R*20R - 20% RAP with rejuvenator

50R -50% RAP with rejuvenator

51

Table 28: ANOVA for Fracture Energy at 25°C for 20% RAP

Table 29: ANOVA for Fracture Energy at 25°C for 50% RAP

Table 28 and Table 29 show the ANOVA results obtained for 20 % RAP and 50 % RAP mixes.

We can see from Table 28 that there is no significant difference between the groups of different

52

mix type. Hence all the samples are ranked ‘A’ in Table 26. From Table 29 also we observe that

there is no significant difference between the groups and they are all ranked as A in Table 27.

Cost analysis

A cost analysis was carried out to understand the relative cost economics of the HMA mixes with

different percentages of RAP with and without rejuvenators. Table 30 shows the relative cost of

each material per ton and the source. It can be seen from this table that major part of the costs

come from the binder in comparison with other materials such as aggregates and RAP. Table 31

shows the relative cost of each rejuvenator based on their density. From Table 32 we can observe

a savings of 40% in total cost when using waste vegetable oil as rejuvenator and 50% RAP in

comparison with virgin mix and a savings of 34% when using Sylvaroad rejuvenator and 50%

RAP in comparison with virgin mix. It can be concluded that a considerable savings in cost can be

accomplished by using higher RAP percentages with rejuvenators. Note that these savings are

based on the percentages of the rejuvenators used in this study

Table 30: Cost of materials

Unit Material Cost ($) Source

per ton Virgin AC 450 MDOT

per ton Aggregate 20 MDOT

per ton RAP 15 Assumed

per gallon WV 3 Frank, 2014 [41]

per gallon SV 9 Arizona Chemicals

Table 31: Density and cost of Rejuvenators

Unit Material Density Cost ($)

lb. per gallon WV, SV 7.6

gallon per lb.

0.13158

gallon per ton 263.158

cost per ton WV

789.47

SV 2368.42

53

Table 32: Comparison of Cost

Unit Mix Cost($)

Savings with

respect to 20%

RAP mix

Savings with respect to all virgin

mix

per

ton 20% RAP 3

per

ton 80% new mix 32

per

ton

Total 20%

RAP Mix 35 20

per

ton 50% RAP 8

per

ton 50% new mix 16

per

ton WV 2

50% RAP mix

with WV 26 25 40

per

ton SV 5

per

ton

50% RAP mix

with SV 29 17 34

per

ton all virgin mix 44

Summary of tests

A radar chart (Figure 18) was plotted with all of the test results to compare the different types of

mixes. We can see from Figure 18 that 50R-SV binder has a higher penetration value compared to

that of 50R-WV. While comparing the complex modulus (G*) and Modulus (Es) of different mixes

we observe that both 50% RAP and the 50% RAP Sylvaroad mix have higher G* and Es indicating

that these mixes have greater resistance to rutting. The 50R-WV has lower G* and also lower ES

compared to the control and 50% RAP indicating a more flexible material even at low

temperatures. The creep compliance test shows that the 50R-SV has higher creep compliance value

compared to the waste vegetable and control mix, indicating a more flexible and durable material

at low temperatures and better resistance to low temperature cracking. The ITS test results show

that the 50% RAP mix has higher tensile strength due the presence of stiff and aged RAP binder

54

in the mix. The rejuvenator mixes have lower ITS value indicating a less stiff material than the

control mix. We can also see considerable savings in cost when using 50R in comparison with

20R. By using rejuvenators the cost increase marginally in comparison with 50R as the cost of

rejuvenators is added. Huge savings in cost can be achieved by using 50% RAP with and without

rejuvenators.

Figure 18: Radar chart of different test results in comparison with control mix

(Values normalized against values of control mix, 20R)

Creep Compliance (1/Gpa)

Pre MIST Es (Mpa)

Post MIST Es (Mpa)

ITS (Mpa)

Back Calculated G* at 25°C (Kpa)

Back Calculated G* at -10°C (Kpa)

Penetration (1/10mm)

Cost ($)

Control (20R) 50R 50R-WV 50R-SV

55

10 CONCLUSIONS

The following conclusions can be drawn from this study:

1. The UPV test results at 25°C indicate that 50% RAP rejuvenated mixes have design moduli

values in between that of 50% RAP and control mix. This indicates the effectiveness of the

rejuvenators in reducing the overall stiffness of the 50% RAP mixes.

2. The MIST results indicate that the recycled mixes with rejuvenator are more prone to

increase in modulus most likely because of increase in density with control mix and 50%

RAP.

3. Binder extracted from 50% RAP mixes with rejuvenators have same or higher PG range

compared to the regularly used binder. The PG sum of the rejuvenated mixes are higher

than the virgin binder.

4. The presence of high recycled asphalt pavement (RAP) content in a mix tends to decrease

the creep compliance and increase the tensile strength compared to a mix with low RAP

content.

5. The addition of the two rejuvenators was found to be effective in increasing the low-

temperature creep compliance value in comparison with the control mix and 50% RAP

mix, therefore implying that rejuvenators are capable of improving low-temperature

performance of RAP mixtures.

6. The SCB test results showed that by increasing the dosage of waste vegetable oil

rejuvenator, the fracture energy of 50% RAP mixes could be reduced and the flexibility

index could be increased, and most likely, resistance against fatigue cracking can be

increased.

7. The savings in cost while using 50% RAP with waste vegetable oil as rejuvenator is around

40% in comparison with virgin mix and while using 50% RAP with Sylvaroad as

rejuvenator it is around 34% in comparison with virgin mix.

8. Overall the properties of 50% RAP mixes with rejuvenator are not inferior in comparison

with 20% RAP mixes.

56

11 RECOMMENDATIONS

The following recommendations are made on the basis of this study:

1. A detailed study with different percentages of rejuvenator (dosages) should be conducted

to determine the most cost effective rejuvenator content.

2. In order to evaluate the relative performance of HMA mixes with 50% RAP, with and

without rejuvenators, test section should be constructed. The performance of these mixes

should be monitored under actual traffic, climate and environmental conditions.

3. The field performance data can be used to develop performance prediction models, which

can be used in the calibration of M-E pavement design equations and life cycle cost can be

determined.

4. Specifications for the use of High RAP mixes with and without rejuvenators may be

developed, so that DoTs can adopt these mixes on a regular basis.

57

12 REFERENCES:

[1]. Indoria R.P., “Nanotechnology in Road construction”, Indian Highways, May 2011.

[2]. Zaumanis Martins, Mallick Rajib, B, Frank Robert. “100% Recycled Hot Mix Asphalt: A

Review and Analysis”, Resources, Conservation and Recycling, 92, 230-245, 2014.

[3]. Kandhal, P. S., and Mallick R. B., “Pavement Recycling Guidelines for State and Local

Governments – Participant’s Reference Book”, Report No. FHWA-SA-98-042, National

Center for Asphalt Technology, Auburn, AL, 1997.

[4]. Huang Baoshan, Li Guoqiang, Vukosaviljevic Dragan, Shu Xiang, Egan Brian K.

“Laboratory investigation of mixing hot-mix asphalt with reclaimed asphalt pavement”

Transportation Research Record 1929. Transportation Research Board (TRB),

Washington, DC, 37–45, 2005

[5]. Zaumanis Martins “100% Recycled Hot Mix Asphalt and use of rejuvenators”, Thesis

Dissertation, 2014.

[6]. Al-Qadi Imad L., Elseifi M. A., and Carpenter Samuel H., “Reclaimed Asphalt Pavement

– A literature Review,” Research Report FHWA-ICT-07-001, Illinois Center for

Transportation, University of Illinois at Urbana-Champaign, Urbana, IL, 2007

[7]. Read, J., Whiteoak, D., & Hunter, R. N., “The Shell Bitumen handbook”. London: Thomas

Telford, 2003.

[8]. Mallick Rajib B and El-Korchi Tahar, “Pavement Engineering: Principles and Practice,

Second Edition, CRC Press, Florida.

[9]. Boyer R.E., "Asphalt Rejuvenators "Fact, or Fable"," Asphalt Institute, 2000

[10]. Zaumanis Martins and Mallick Rajib B. “Finite Element Modeling of Rejuvenator

Diffusion in RAP Binder Film – Simulation of Plant Mixing Process” In: Proceedings of

the International RILEM Symposium Stockholm, Stockholm, Sweden, June 10–12, 407–

19, 2013.

[11]. Hansen K. R. and Copeland A., “2nd Annual Asphalt Pavement Industry Survey on

Reclaimed Asphalt Pavement, Reclaimed Asphalt Shingles, and Warm-Mix Asphalt

Usage: 2009-2011," National Asphalt Pavement Association, Lanham, MD, 2013

[12]. Copeland A., "High Reclaimed Apshlat Pavement Use. Info Brief," Federal Highway

Administration, Washington, D.C., 2011.

58

[13]. Chesner WH, Collins RJ, MacKay MH. “Users guidelines for waste and byproduct

materials in pavement construction”, Report No. FHWA-RD-97-148, Commack: Chesner

Engineering, P.C., April 1998

[14]. Thomas W. Kennedy, Weng O. Tam, and Mansour Solaimanian “Effect of reclaimed

asphalt pavement on binder properties using the Superpave system”, Center for

transportation research bureau of engineering research the University of Texas at Austin,

September 1998.

[15]. Bennert Thomas, Daniel Jo Sias and Mogawer Walaa., “Strategies for Incorporating

Higher Recycled Asphalt Pavement percentages – Review of Implementation Trials in

Northeast States”, Journal of the Transportation Research Board No. 2445, 83-93., 2014.

[16]. Chen Jian-Shiuh, Chen Shih-Fan, Liao Min-Chih and Huang Shou Wei.”Laboratory

Evaluation of Asphalt Blends of Recycling Agents with Aged Binders”, J.

Mater.Civ.Eng., 27(4), 2015.

[17]. Shen Junan, Amirkhanian Serji and Miller Jenifer Aune., “Effects of Rejuvenating Agents

on Superpave Mixtures Containing Reclaimed Asphalt Pavement”, J. Mater.Civ.Eng.,

9(5), 376-384, 2007

[18]. Li Jian, Ni Fujian, Huang Yemin and Gao Lei., “New Additive for Use in Hot-In-Place

Recycling to Improve Performance of Reclaimed Asphalt Pavement Mix”, Journal of the

Transportation Research Board, No.2445, 39-46, 2014

[19]. Bennert Thomas, Daniel Jo Sias and Mogawer Walaa. (2014), “ Strategies for

Incorporating Higher Recycled Asphalt Pavement percentages – Review of

Implementation Trials in Northeast States”, Journal of the Transportation Research Board

No. 2445, 83-93

[20]. Cooper Samuel B Jr, Mohammad Louay N, Elseifi Mostafa and Medeiros Marcelo, Jr., “

Effect of Recycling Agents on the Laboratory Performance of Asphalt Mixtures

Containing Recycled Asphalt Shingles”, Journal of the Transportation Research Board

No. 2506, 54-61., 2015

[21]. Mogawer Walaa S, Austerman Alexander J, Kluttz Robert and Puchalski Sabastian,

“Using Polymer Modification and Rejuvenators to Improve the Performance of High

Reclaimed Asphalt Pavement Mixtures”, Journal of the Transportation Research Board

No.2575, 10-18, 2016

59

[22]. Diefenderfer Brian K, Bowers Benjamin F, Schwartz Charles W, Farzaneh Azadeh abd

Zhang Zhuoyi., “ Dynamic Modulus of Recycled Asphalt Mixtures, Journal of the

Transportation Research Board No.2575, 19-26., 2016

[23]. Ding Yongjie, Huang Baoshan, Shu Xiang, Zhang Yuzhen and Woods Mark, E ., “Use

of Molecular Dynamics to Investigate Diffusion Between Virgin and Aged Asphalt

Binders”, Fuel, 174, 267-273, 2016

[24]. Shen Junan, Amirkhanian Serji, Tang Boming, “Effect of Rejuvenator on Performance

Based Properties of Rejuvenated Asphalt Binder and Mixtures”, Construction and

Building Materials, 21, 958-964. 2007.

[25]. Im Soohyok, Zhou Fujie, Lee Robert and Scullion Tom, “Impacts of Rejuvenators on

Performance of Engineering Properties of Asphalt Mixtures Containing Recycled

Materials”, Construction and Building Materials, 53, 596-603., 2014.

[26]. Zaumanis Martins, Mallick Rajib B and Frank Robert., “ Determining Optimum

Rejuvenator Dose for Aspahlt Recycling Based on Superpave Performance Grade

Specifications”, Construction and Building Materials, 69, 159-166., 2014.

[27]. Zaumanis Martins, Mallick Rajib B, Poulikakos Lily and Frank Robert, “Influence of Six

Rejuvenators on the Performance Properties of Reclaimed Asphalt Pavement (RAP)

Binder and 100% Recycled Asphalt Mixtures”, Construction and Building Materials, 71,

538-550., 2014.

[28]. Buss Ashley, Williams Christopher, R and Schram Scott., “The Influence of Warm Mix

Asphalt on Binders in Mixes That Contain Recycled Asphalt Materials”, Construction and

Building Materials, 77, 50-58., 2015.

[29]. Ongel Aybike, Hugener Martin, “Impact of Rejuvenators on aging Properties of

Bitumen”, Construction and Building Materials, 94, 467-474, 2015.

[30]. Xiao Feipeng, Putman Bradley, Amirkhanian Serji., “Rheological Characteristics

Investigation of High Percentage RAP binders with WMA Technology at Various Aging

States”, Construction and Building Materials, 98, 315-324., 2015

[31]. Nazzal Munir D, Mogawer Walaa, Austerman Alexander, Qtaish Lina Abu, Kaya Savas,

“ Multi-scale Evaluation of the Effect of Rejuvenators on the Performance of High RAP

Mixtures”, Construction and Building Materials, 101, 50-56., 2015

60

[32]. Ma Tao, Wnag Hao, Huang Xiaoming, Wang Zilong, Xiao Feipeng., “ Laboratort

Performance Characteristics of High Modulus Asphalt Mixture with High – Content

RAP”, Construction and Building Materials, 101, 975-982., 2015

[33]. Ali Ayman W, Mehta Yosuf A, Nolan Aaron, Purdy Caitlin, Bennet Thomas.,

“Investigation of the Impacts of Aging and RAP percentages on Effectiveness of Asphalt

Binder Rejuvenators”, Construction and Building Materials, 110, 211-217., 2016.

[34]. Lu Dai Xuan, Saleh., “ Laboratory Evaluation of Warm Mix Asphalt Incorporating High

RAP Proportion by Uisng Evotherm and Sylvaroad Additives”, Construction and Building

Materials, 114, 580-587., 2016.

[35]. Moghaddam Taher Baghaee, Baaj Hassan., ‘The Use of Rejuvenating Agents in

Production of Reccled Hot Mix Asphalt: A Systematic Review”, Construction and

Building Materials, 114, 805-816., 2016.

[36]. Yu, Xiaokong, Martins Zaumanis, Salomé Dos Santos, and Lily D. Poulikakos.

"Rheological, Microscopic, and Chemical Characterization of the Rejuvenating Effect on

Asphalt Binders." Fuel 135 (2014): 162-71

[37]. Rojas, J., Nazarian, S., Tandon, V., and Yuan, D., Quality management of asphaltconcrete

layers using wave propagation techniques. University of Texas at El Paso. 1999.

[38]. Kakar Mohamad Rafiq, Hamzah Meor Otheman, Valentin Jan., “A Review on Moisture

Damages of Hot and warm Mix Asphalt and Related Investigations”, Journal of Cleaner

Production, 99, 39-58. 2015.

[39]. Tran, N., A. Taylor, and J.R. Willis, “Effect of Rejuvenator on Performance Properties of

HMA Mixtures with High RAP and RAS Contents”, NCAT Report 12-05, National

Center for Asphalt Technology, June 2012.

[40]. Willis, J.R., P. Turner, G. Julian, A.J. Taylor, N. Tran, and F.D.G. Padula, “Effects of

Changing Virgin Binder Grade and Content on RAP Mixture Properties”, NCAT Report

12- 03, National Center for Asphalt Technology, May 2012.

[41]. Frank, Bob. Personal communication; June 25, 2012

[42]. Braham, A. F., W. G. Buttlar, T. R. Clyne, M. O. Marasteanu, and M. I. Turos. “The

Effect of Long-Term Laboratory Aging on Asphalt Concrete Fracture Energy”. Journal of

the Association of Asphalt Paving Technologists, Vol. 78, pp-417-454, 2009.

61

[43]. Nazarian S, Yuan D, Tandon V, Arellano M. “Quality management of flexible pavement

layers with seismic method”. Research report 1735-3, Center for Transportation

Infrastructure Systems, University of Texas at El Paso, 2002

[44]. Palisade Corporation 2015 (The DecisionTools Suite), New York.


Recommended